Categorization Of Land Area Using K-Means And Fuzzy C-Means Clustering Algorithms

Document Type : Primary Research paper

Authors

1 Bannari Amman Institute of Technology, Sathyamangalam, TamilNadu

2 Vishveshwaraiah Institute of Science & Technology, Chittoor, Andhra Pradesh.

Abstract

Division and request of significant standard satellite imagery is a troublesome
issue because it's important to finish this errand on a pixel-by-pixel premise. The fine spatial
aim implies that each item is made up of a collection of pixels in close spatial proximity
and precise order, necessitates that this perspective be considered without delay. For arranging
high aim satellite symbolism, K-implies grouping calculation is a superior technique.
k-implies, also known as Lloyd's computation, is an iterative information parceling
calculation that assigns perceptions to one of several clusters defined by centroids, with k
chosen before the algorithm begins. A basic distance choosing principle is used to arrange
the eliminated areas. The technique effectively eliminates the blended pixel problem that
plagues most pixel-based techniques. To group satellite symbolism into explicit articles
within its boundaries and for ecological organizing reasons, we used K-implies and fuzzy C
techniques grouping calculation. Clusters are formed from the various pixels. For each
centroid, a center point is discovered. This centroid is used to classify the entire cluster. In
this research, the k-means clustering technique and fuzzy c-means were used to separate
the pixels of satellite images and process them using MATLAB software before naming the
clusters based on their color configurations. This clustering algorithm was used to colorcode
the locations in the satellite image. For example, green denotes a forest, blue denotes
a body of water, brown denotes a muddy terrain, and so on. Fuzzy C-means is independent
of the original groupings, resulting in improved clustering outcomes.

Keywords